知識本體(Ontology)是一種用來表達知識的方式,其被廣泛的應用於各種領域中。知識本體利用簡潔的方式表示該領域中所含有的知識概念以及知識概念相互之間的關係,讓知識本體可以被電腦解讀並加以使用。然而目前在建置知識本體的流程上,並未有統一的標準與方法,且在建置知識本體的過程中,需要大量領域專家的投入,並隨著領域中知識的發展,必須不斷對知識本體進行增補。因此如何以自我學習的方式來建議知識本體的增補內容,以減少領域專家人力的參與,是建置知識本體時的重要議題。 本研究的主要目的為縮短增補特定領域的知識本體內容時,所需花費的時間與人力。為此在本研究中選定「建築資訊塑模(Building Information Modeling)」領域作為研究的特定領域,由分析塑模領域的文獻開始,收集文獻中所整理的知識概念以建置塑模領域的基本知識本體,並建立該領域的文件集合,以驗證使用知識本體的資訊檢索技術之成效表現乃優於使用向量空間模型的資訊檢索技術,再針對資訊檢索的檢索結果進行知識概念擷取,來對知識本體的內容進行補充。本論文將詳述各步驟所進行之研究內容,並先透過實驗來驗證應用知識本體進行資訊檢索確實能提高資訊檢索的成效表現,再驗證經過資訊檢索技術所增補的知識本體亦能提高資訊檢索之成效。
Ontology, which has been widely used in different domains, concisely represents the knowledge as a set of concepts and the relations of those concepts. However, with the growth of the domain knowledge and its lack of unified standards, building and revising the ontology is not only time-consuming but also requires a large amount of manpower. For making the process more efficient, this research proposed a self-learning method to suggest the enhancement of the ontology on a specific domain. In this research, the specific domain is focused on building information modeling (BIM). There are three steps of the research progress. First, collect concepts from the related researches as reference to build the base ontology. Then, propose an ontology-based retrieval model to improve the retrieval effectiveness. Finally, propose a methodology to extract the concepts from the ontology-based retrieval results. According to the experiment results, using the enhanced ontology to the ontology-based retrieval could improve the retrieval effectiveness. The enhanced ontology can also help learning and sharing the domain knowledge.
為了持續優化網站功能與使用者體驗,本網站將Cookies分析技術用於網站營運、分析和個人化服務之目的。
若您繼續瀏覽本網站,即表示您同意本網站使用Cookies。